Specific-site methylation of tumour suppressor ANKRD11 in breast cancer

European Journal of Cancer (2012) 48, 3300–3309

Available at www.sciencedirect.com

journal homepage: www.ejcancer.info

Specific-site methylation of tumour suppressor ANKRD11 in breast cancer Sue Ping Lim a,⇑, Nick C. Wong b, Rachel J. Suetani a, Kristen Ho a, Jane Lee Ng b, Paul M. Neilsen a, Peter G. Gill c, Raman Kumar a, David F. Callen a a

Cancer Therapeutics Laboratory, Discipline of Medicine, The University of Adelaide, South Australia, Australia Cancer and Disease Epigenetics Laboratory, Murdoch Childrens Research Institute, Royal Children’s Hospital, Department of Paediatrics, University of Melbourne, Parkville, Victoria, Australia c Discipline of Surgery, Royal Adelaide Hospital, The University of Adelaide, South Australia, Australia b

Available online 24 April 2012

KEYWORDS Specific DNA methylation ANKRD11 Breast cancer Tumour suppressor Promoter activity Decitabine Zebularine SEQUENOM

ANKRD11 is a putative tumour suppressor gene in breast cancer, which has been shown in our laboratory to be a co-activator of p53. Our data suggest that down-regulation of ANKRD11 is associated with breast tumourigenesis. Breast cancer cell lines treated with DNA demethylating agents resulted in up-regulation of ANKRD11 expression suggesting that promoter DNA methylation may be responsible for its down-regulation. The transcriptional activity of a CpG-rich region 2 kb upstream of the transcription initiation site of ANKRD11 was investigated using dual-luciferase reporter assays. The constructs carrying –661 to –571 bp promoter sequence showed significant transcriptional activity. Using the SEQUENOM Epityper Platform, the region between –770 and +399 bp was analysed in 25 breast tumours, four normal breast tissues and five normal blood samples. The region between –770 and –323 bp was shown to be frequently methylated in breast tumours. The methylation patterns of all analysed CpGs in this region were identical in the normal and tumour samples, except for a 19 bp region containing three CpG sites. These sites had significantly higher levels of methylation in tumours (40%) compared to normal samples (6%). Our findings support the role of ANKRD11 as a tumour suppressor gene and suggest that aberrant DNA methylation of three CpGs in a 19 bp region within the ANKRD11 promoter may be responsible for its downregulation in breast cancer. Crown Copyright Ó 2012 Published by Elsevier Ltd. All rights reserved.

Abstract

1. Introduction ⇑ Corresponding author: Address: Cancer Therapeutics Laboratory, Discipline of Medicine, University of Adelaide, Frome Road, Adelaide, South Australia 5000, Australia. E-mail address: [email protected] (S.P. Lim).

Breast cancer is one of the most common cancers in women, and in 2008, it accounted for 6% of cancerrelated deaths worldwide.1 Cancer is driven by activation

0959-8049/$ - see front matter Crown Copyright Ó 2012 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejca.2012.03.023

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

of oncogenes and inactivation of tumour suppressor genes. Inactivation of tumour suppressor genes results from mutation, loss of heterozygosity (LOH) and/or epigenetic silencing.2 ANKRD11/ANCO-1 is a putative breast cancer suppressor and is a co-activator of p53.3 ANKRD11 is also shown to interact with p160 nuclear receptor co-activators and inhibits ligand-dependent transcriptional activation.4 Recent reports suggest that ANKRD11 is a candidate gene for autism and neurocognitive impairments in patients with 16q24.3 microdeletion syndrome also suggest its normal role in development.5,6 ANKRD11 is located at 16q24.3, a predominant region of LOH in breast cancer.7 LOH is a common mechanism for loss of one copy of a tumour suppressor gene. The other copy can be inactivated by an epigenetic mechanism, such as mutation or DNA methylation, resulting in reduced ANKRD11 expression. DNA methylation is catalysed by DNA methyltransferases (DNMTs), where the recruitment of methyl-binding proteins and histone deacetylases results in the formation of transcriptionally repressive chromatin states.8 Typically, gene transcription is silenced by hyper-methylation of the CpG-rich promoter region, which is usually located in the 50 untranslated region (50 UTR) or the 50 exon/intron region of a gene.9,10 Herein, we determined the expression of ANKRD11 in both normal and breast cancer tissues and investigated the role of CpG methylation of the promoter of ANKRD11 in modulating its expression. Our results suggest that specific CpG methylation of ANKRD11 promoter is associated with breast cancer. 2. Materials and methods 2.1. Clinical sample collections Tissues were obtained with informed consent from 30 breast cancer patients and five normal breast reduction mammoplasties performed at the Royal Adelaide Hospital (RAH) between 2003 and 2011. Formalin fixed paraffin-embedded tissue was used for immunohistochemistry. Unfixed tissues were stored in RNAlater solution (Ambion) at –20 °C and subsequently used for DNA and RNA extraction using the Allprep DNA/RNA mini kit (Qiagen). Relevant clinical data were retrieved from patient’s records including human epidermal growth factor receptor 2 (HER2), oestrogen receptor (ER), progesterone receptor (PR) and proliferation index (MIB-1) status. Genomic DNA extracted from five normal blood samples was kindly supplied by Dr. Kathryn Friend of the Women’s and Children’s Hospital (Adelaide). 2.2. Cell culture All human breast cell lines were purchased from the American Type Culture Collection (ATCC). MCF-7

3301

and ZR75-1 cells were maintained in RPMI supplemented with 10 lg/mL of insulin (Invitrogen) and 1 mM of sodium pyruvate. MDA-MB-231 and MDAMB-468 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM). Media were supplemented with 10% foetal bovine serum and 100 U/mL of penicillin–streptomycin–glutamine (Invitrogen). MCF-10A cells were cultured in DMEM nutrient mixture F12 HAM with 20 ng/mL EGF, 0.5 lg/mL of hydrocortisone, 100 ng/mL of cholera toxin, 10 lg/mL of insulin and 5% horse serum (Invitrogen). All reagents were purchased from Sigma unless specified. Cells were grown at 37 °C with 5% CO2. 2.3. Treatment with DNMT inhibitors Twenty-four hours prior to treatment, cells were plated at 1  105 cells/well of a 6-well plate. Treatments consisted of 5-aza-20 -deoxycytidine (decitabine; Sigma) or 2(1H)-pyrimidinone riboside (zebularine; Sigma) for 72 h. Drug levels were maintained by replacing media containing the relevant concentration of drug every 24 h. Following completion of the experiment, cells were harvested and RNA and DNA extracted using the RNeasy plant mini kit (Qiagen) and the DNeasy Blood and tissue kit (Qiagen), respectively. 2.4. Luciferase reporter system 2.4.1. ANKRD11 promoter activity and site-directed mutagenesis Six different regions of the human ANKRD11 gene including 50 UTR and exon 1 (–2000 to +306) were cloned into a promoterless luciferase reporter vector, pGL3-Basic (Promega, Madison, WI), designated as P1 (–2000 to +306), P2 (–2000 to –661), P3 (–571 to +306), P4 (–862 to +107), P5 (–689 to +306) and P6 (–689 to –543) (Figs. 3 and 5). Specific mutations in the P6 construct (Fig. 5B) were generated by overlap PCR using primers listed in Table S2. All constructs were verified by DNA sequencing. 2.4.2. In vitro methylation of ANKRD11 promoter Recombinant HpaII, M.SssI and HhaI methylases (New England Biolabs) were used to methylate specific sites of the pGL3-P6 construct. Plasmid DNA (4 lg) was treated with 4 units of the specific methylase and 640 lM of S-adenosylmethionine in the manufacturer’s recommended buffer at 37 °C for 4 h, incubated at 65 °C for 20 min to inactivate the enzymes and then purified using QIAquick PCR kit (Qiagen). Methylation was confirmed by the lack of endonuclease restriction site by HpaII for HpaII methylation; BstUI for M.SssI methylation; and HinP1I for HhaI methylation.

3302

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

2.4.3. Transient transfection and luciferase assay Briefly, 1  105 of MCF-10A cells were plated/well of 24-well tissue culture dishes. Transfections were performed with lipofectamine 2000 reagent (Invitrogen) following the manufacturer’s protocol. An internal control plasmid for Renilla luciferase expression, pRL-TK (Promega), was included in each transfection. Twenty-four hours post-transfection, cells were lysed and the Firefly and Renilla luciferase activities were sequentially measured using the Dual-Luciferase Reporter Assay System (Promega) following the manufacturer’s instructions. The relative luciferase activity was calculated by the ratio of firefly and Renilla expression in each sample. All transfections were carried out in at least two independent experiments in triplicate. Data are expressed as mean values with standard errors, and were analysed by unpaired two-tailed t-test. 2.5. Quantitative real-time polymerase chain reaction (qRT-PCR) Reverse transcription was performed using 1 lg of total RNA, random primers and 20 U of reverse transcriptase (Promega) in a total volume of 50 lL. ANKRD11, DNMT1 and DNMT3B expressions were determined by qRT-PCR using IQe SYBR green supermix (Biorad) and primers listed in Table S1. Cycling conditions were: 10 min at 95 °C followed by 40 repeats of the following cycle: 95 °C for 10 s, annealing at the appropriate temperature for 10 s and extension at 72 °C for 10 s. b-Actin expression was used for normalisation of target gene expression. 2.6. SEQUENOM MassARRAY Epityper for methylation analysis DNA samples were bisulphite-treated and purified using the Epitect Kit (Qiagen) according to the manufacturer’s protocol. Unconverted genomic DNA was included as a negative control. Primer design ensured the PCR amplification of completely converted DNA. Three overlapping regions (PF1, PF2 and PF3) of the bisulphite-treated ANKRD11 promoter sequence were amplified using the PyroMark Kit (Qiagen) and primers in Table S1. As described previously, in vitro transcription and uracil-specific cleavage of the amplified products were undertaken before analysis of the samples by matrix-assisted laser desorption and ionisation time-offlight mass spectrometry (MALDI-TOF-MS).11 3. Results 3.1. ANKRD11 is down-regulated in human breast tissues We assayed the levels of ANKRD11 mRNA in human invasive breast tumour tissues and normal samples using

qRT-PCR (Fig. 1A). All but one of the tumour samples (T14) had a lower level of mRNA (3.9 ± 0.8, n = 30) than the normal breast tissues (15.5 ± 7.5, n = 5) (p < 0.01, Fig. 1A and B). Interestingly, the tumour sample with higher levels of ANKRD11 expression was a papillary breast cancer (T14, Table 1). Notably, the primers used to assay levels of ANKRD11 mRNA would bind both variants a [XR_123180.1] and b [NM_013275.4] that share common mRNA sequences at the 30 end. To determine if one of these variants was responsible for higher ANKRD11 expression in T14, variant a-specific primers were designed. The result showed that the variant b [NM_013275.4] of ANKRD11 was responsible for the higher expression in the single case of papillary carcinoma, T14 (Fig. S1). 3.2. ANKRD11 expression in breast cell lines is restored by DNMTs inhibitors We next measured levels of ANKRD11 mRNA in breast cancer cell lines. MDA-MB-231 (ER–), MCF-7 (ER+), MDA-MB-468 (ER–), ZR75-1 (ER+) and the immortalised breast cell line MCF10A (ER–) have low levels of ANKRD11 expression when compared with normal breast tissue (N1, N2; Fig. 2A). To investigate whether ANKRD11 inactivation in breast cell lines is caused by DNA methylation, we treated these cell lines with demethylating agents (DNMT inhibitors).12 The breast cell lines showed variable response to two DNMT inhibitors, decitabine and zebularine. Modest increases in levels of ANKRD11 expression were observed in decitabine treated MCF-10A and MDA-MB-468 (Fig. 2B and C). MDA-MB-231, MCF-7 and ZR75-1 showed no response to decitabine treatments (Fig. 2D–F). Zebularine was the most effective agent as it induced at least 5-fold increases in ANKRD11 expression at the highest concentration used (500 lM), in all cell lines except MCF-7 (Fig. 2B–F). These findings suggest that zebularine and decitabine responses are cell-type dependent. Taken together, these results suggest that ANKRD11 could be down-regulated by DNA methylation as ANKRD11 gene activity was induced in response to DNMT inhibitors. 3.3. Identification of ANKRD11 promoter Since demethylating agents can activate the expression of ANKRD11 in breast cell lines, promoter activity of this gene was characterised in order to further study its methylation status. The ANKRD11 gene contains 13 exons and is transcribed from within exon 3 (Fig. 3A). A CpG island was identified that extends from exon 1 to 800 bp upstream of transcription start site. A 2 kb region identified as a putative promoter region was tested for its transcriptional activity using a dual-luciferase reporter assay (Fig. 3B). A 2317 bp

3303

50 40 30 20 10 0

B

Relative ANKRD11 expression

T1 T2 T3 T4 T5 T6 T7 T8 T9 T1 T10 T11 T12 T13 T14 T15 T16 T17 T18 T29 T20 T21 T22 T23 T24 T25 T26 T27 T28 T39 0 N 1 N 2 N 3 N 4 N 5

A

Relative ANKRD11 expression

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

** 50 40 30 20 10 0 Tumour

Normal

Fig. 1. ANKRD11 is down regulated in breast tumours. (A) The ANKRD11 mRNA levels in tumour (T) and normal (N) samples normalised to housekeeping gene b-actin. (B) The average mRNA levels of all tumour (n = 30) and normal samples (n = 5). An unpaired two-tailed t-test was used. The data are presented as mean ± SEM; **p < 0.005.

Table 1 Clinical data of human breast samples used in this study. Samples

ER

PR

HER2

MIB-1 count (%)

Tumour type

Grade

T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13 T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28 T29 T30

+ – + – – + + + – – + + + + + + + + + + + + – + + + + – – +

+ – + – – – + + – – + + – + + – + + + + + + + + + + + – – +

– – ++ – – – +++ – – +++ +++ ++ – – +++ – + – + – – +++ – + ++ + – ++ +++ +++

<20 >30 >30 >30 >30 >30 >30 >30 >30 >30 >30 20–30 <20 20–30 <20 <20 <20 <20 <20 >30 <20 >30 <20 <20 20–30 <20 >30 >30 >30 >30

IDC/ILC IDC IDC IDC IDC ILC IDC IDC IDC IDC IDC IDC IDC IPC IDC ILC IDC IDC IDC IDC ILC IDC ILC IDC IDC IDC ILC IDC IDC IDC

1 2 3 2 3 2 3 2 3 3 3 2 1 1 2 2 1 1 3 2 1 3 3 1 2 2/3 2 3 3 3

ER, PR, HER2, MIB-1 count, tumour type and grades of each tumour sample are shown.

3304

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

Breast cell lines

B

Human breast tissues Relative ANKRD11 expression

150

100

50

20 15

****

10

*

5 0

T3 N 1 N 2

T2

MCF-10A

25

DMSO

Decitabine

*** ****

Zebularine

M

T1

0

B 23 M 1 C F M 7 B 46 ZR 8 7 M 5-1 C F1 0A

Relative ANKRD11 expression

A

Concentration

D

25

MDA-MB-231

20 15 10

****

5 0

*

* DMSO

Decitabine

Zebularine

Relative ANKRD11 expression

Relative ANKRD11 expression

C

25

MDA-MB-468

20 15 10

***

5

* 0

DMSO

Concentration

F

25

MCF-7

20 15 10 5 DMSO

Decitabine

Zebularine

Concentration

Relative ANKRD11 expression

Relative ANKRD11 expression

Zebularine

Concentration

E

0

Decitabine

**** ****

25

ZR75-1

**

20 15 10 5 0

DMSO

Decitabine

Zebularine

Concentration

Fig. 2. ANKRD11 is up-regulated in breast cell lines after treatment with DNA methyltransferase (DNMT) inhibitors. (A) The relative expression of ANKRD11 was measured in breast cell lines and human breast tissues as described in Fig. 1A. ANKRD11 mRNA level is shown relative to vehicle control in (B) the immortalised non-malignant breast cell line, MCF-10A and, breast cancer cell lines: (C) MDA-MB-468, (D) MDA-MB231, (E) MCF-7 and (F) ZR75-1. The cells were treated with 1, 10, 50 lM decitabine or 50, 100, 500 lM zebularine for 72 h. The relative ANKRD11 expression was normalised to b-actin. Unpaired two-tailed t-tests for all statistical analysis compared to vehicle control. *p < 0.05, **p < 0.005, *** p < 0.001.

fragment of ANKRD11, including exon 1 and 2 kb upstream (designated as P1), was cloned into a luciferase reporter vector, together with various sub-fragments of P1 (Fig. 3B). The constructs P4 (–862 to +107) and P5 (–689 to +306) showed the highest promoter activity. However, constructs P1 and P3 exhibited modest levels of transcriptional activity that were completely abolished in P2, which lacked the –660 to +308 bp region. These results suggest that sequences regulating the

promoter activity of ANKRD11 are located within the –689 to +107 bp region (Fig. 3B). 3.4. The ANKRD11 promoter is hypermethylated in breast cancer Following identification of a putative ANKRD11 promoter region, the total methylation status of three overlapping regions of the promoter (PF1 to PF3) was

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

3305

A

B -2000

0

+306

CpG

pGL3 -2000

+306

pGL3-

-661

LUC

P2

LUC

P3 -862

+107

LUC

P4 +306

50

40

30

LUC

20

P5

10

-689

0

pGL3

+306

-571

pGL3pGL3-

LUC

P1

-2000

pGL3-

LUC

Relative of luciferase activity Fig. 3. Characterisation of the ANKRD11 promoter. (A) Genomic organisation of ANKRD11. The 223 kb ANKRD11 gene is transcribed to a 9.3 kb mRNA. (B) The transcriptional activity was measured for regions of ANKRD11 upstream of the transcriptional initiation site. CpG islands are shown as vertical bars. Promoter constructs P2, P3, P4 and P5 represent different parts of P1. Values represent averages of triplicate treatments from at least two independent experiments. **p < 0.005.

analysed in breast tumour and normal breast samples using MALDI-TOF-MS (Fig. 4A). SEQUENOM MassARRAY Epityper employs base-specific cleavage followed by MALDI-TOF-MS to quantify DNA methylation of CpG sites within a target region. DNA methylation of CpG sites across each amplicon, PF1 (–770 to –323 bp), PF2 (–299 to +15 bp) and PF3 (–23 to +399 bp), were averaged to identify the amplicon with higher methylation in breast tumour samples. A significantly higher methylation was observed in the PF1 region compared with regions PF2 and PF3 (Fig. 4B). This suggests that the promoter region between –770 and –323 bp contains methylated CpGs that may be responsible for the aberrant ANKRD11 expression in breast cancer.

3.5. Site-specific methylation of ANKRD11 promoter Since the PF1 region was the significantly methylated part of the ANKRD11 promoter, 17 fragments (units) of PF1 were analysed by mass spectrometry (Fig. 4C). A comparison of DNA methylation status in seven normal and 25 breast cancer samples revealed that differences in methylation between normal and tumour breast tissues were localised to CpG unit 7. CpG unit 7 of PF1 is a 19 bp fragment that includes three CpG sites (–582 to

–574 bp). This unit was highly methylated in 36% (9/ 25) of the tumour samples analysed (Fig. 4D). We also investigated if there was any relationship between ANKRD11 methylation status and clinical markers in breast cancer samples. The clinical data showing ER, PR, HER2 and MIB-1 status of the breast tumours are given in Table 1. Methylation at CpG unit 7 was observed to be significantly associated with HER2 negativity (p < 0.05) (Fig. 4E), but was not associated with tumour grade, ER, PR or MIB-1 status (Fig. S2). In human tumour samples, there was a trend for high levels of methylation at CpG unit 7 to be associated with lower levels of ANKRD11 expression (Fig. 4F). However, some tumour samples displayed low ANKRD11 expression in the absence of elevated methylation, suggesting that in these samples ANKRD11 is down-regulated by alternative mechanism. A similar trend was also seen in the breast cell lines. For example, the three cell lines with the higher level of methylation at CpG unit 7, MDA-MB-468, MDAMB-231 and MCF10A, have a relatively low level of ANKRD11 expression (Fig. 4G). These cell lines showed a reduction in methylation at CpG unit 7 after treatment with 500 lM of zebularine (Fig. 4H), which was previously shown in Fig. 2 as the most effective dose. No reduction was observed in ZR75-1 and MCF-7 cells

3306

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

B

A

* ***

-770

PF1

Methylation %

100

-323 -299

15

PF2 -23

80 60 40 20

399

PF3

0 PF1

-770

C

PF3

-323

PF1

100 Methylation % (PF1)

PF2

80 *

Unit 7

60

-582

-574

CpG sites 23,24,25

40 20 0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

CpG units

D Tumour

Methylation % (Unit 7)

100

Normal

80 60 40 20

T1 T2 T3 T4 T6 T7 T8 T9 T1 0 T1 1 T1 2 T1 3 T1 4 T1 5 T1 6 T1 7 T1 9 T2 0 T2 2 T2 4 T2 5 T2 6 T2 7 T2 8 T2 9 N 1 N 2 B 1 B 2 B 3 B 4 B 5

0

60 40 20

Positive

HER2 status

Negative

Human breast tissues 10

G

Breast cell lines 100

Methylation % (Unit 7)

Methylation % (Unit 7)

80

0

H

F

* 100

Relative ANKRD11 expression

E

8 6 4 2 0

MB468

80

MB231 MCF10A

60 R 2= -0.71 p=0.07

40 20 MCF7

0 High Low Methylation level at unit 7

0

1 2 3 4 Relative ANKRD11 expression

ZR75-1

5

PF1 -582

-574

MCF10A MCF10A_Zeb MB231 MB231_Zeb MB468 MB468_Zeb MCF7 MCF7_Zeb ZR75-1 ZR75-1_Zeb Water control

Fig. 4. Methylation patterns of the ANKRD11 promoter in human breast cancer. (A) Location and (B) methylation analysis of three regions (PF1, PF2 and PF3) within the ANKRD11 promoter in breast tumours. Each bar represents the average methylation level of all CpG units in each amplicon (PF1 = 17 units, PF2 = 15 units, PF3 = 10 units). (C) Site-specific methylation levels of ANKRD11 promoter region PF1. The % methylation of tumours (n = 25) and normal samples (n = 7) was averaged in each CpG units. The CpG units are as defined in supplementary Fig. S4. The methylation pattern of PF2 and PF3 are shown in supplementary Fig. S5. (D) Specific methylation levels at CpG unit 7 of PF1 in tumour and normal samples. Tumours (black), normal (white) and blood samples (grey). T5, T18, T21, T23, T30, N3, N4 and N5 are not shown. (E) The relationship of human epidermal growth factor receptor 2 (HER2) status and methylation level at unit 7. (F) Promoter methylation of ANKRD11 promoter at CpG unit 7 and ANKRD11 mRNA levels in human breast tissues. The data from T14 (a different type of breast tumour) and N2 were treated as outliers. High = samples with >50% methylation at unit 7; low = <50% methylation at unit 7. (G) Correlation of promoter methylation of ANKRD11 at unit 7 and relative ANKRD11 expression and (H) PF1 epigram of breast cell lines treated with 500 lM of zebularine (Zeb) for 72 h. Bracket: three CpGs in unit 7. An unpaired two-tailed t-test was used for all statistical analysis. *p < 0.05. ***p < 0.001.

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

due to the low initial methylation level at this locus (Fig. 4H). Overall, these data suggested a relationship between specific methylation at CpG unit 7 of the ANKRD11 promoter and total ANKRD11 expression. 3.6. Site-specific mutation and DNA methylation silence ANKRD11 expression To examine whether the activity of the three CpGs in the unit 7 methylation-sensitive region (–582 to –574 bp) regulates ANKRD11 expression, specific CpGs in the P6 construct (–689 to –543 bp) were either methylated by specific enzymes (Fig. 5A) or mutated by site-specific mutagenesis (Fig. 5B). In in vitro methylation study, methylase enzymes were used to methylate specific CpG sites in the P6 promoter fragment. Promoter activities of each fragment were then assessed. The transcriptional activities of all in vitro P6 constructs were significantly reduced after being methylated (Fig. 5A). Methylation of the third CpG in unit 7 by HpaII methylase more strongly repressed the promoter activity compared to the HhaI methylated P6, which methylated the CpG adjacent to unit 7. Site-directed mutagenesis of the first two CpGs and all three CpGs in unit 7 significantly reduced transcriptional activity (Fig. 5B). However, mutation of the third CpG alone did not show any relationship in transcriptional regulation of ANKRD11, suggesting the critical region was 10 bp region consisting the first two CpGs. 3.7. Relationship of DNMTs and ANKRD11 expression DNA methyltransferase 1 (DNMT1) and DNA methyltransferase 3B (DNMT3B) are responsible for the maintenance and establishment of DNA methylation patterns, respectively.13 DNMT3B is frequently over-expressed in breast tumours.14,15 To identify whether DNMT1 and DNMT3B have roles in regulating ANKRD11 expression, mRNA levels of DNMT1 and DNMT3B were measured in the breast tumour and normal samples. In our dataset, DNMT3B was over-expressed in 30% of the tumour samples and was very high in higher grade tumours (Fig. S3).15 However, there was no relationship between the DNMT3B and ANKRD11 expressions (Fig. S3). In addition, DNMT1 expression varied among breast tumours, normal breast tissues and normal blood samples (Fig. S3) consistent with the published data.15 4. Discussion In this study, mRNA expression of ANKRD11 was shown to be uniformly down-regulated in 23 cases of infiltrating ductal, five cases of lobular and a case of mixed ductal and lobular breast cancer tissues compared to normal breast tissues. The exception was a single case

3307

of invasive intracystic papillary cancer, a breast cancer subtype that has a good prognosis.16 Additional studies are needed to further investigate the possible relationship of ANKRD11 expression and papillary breast cancer. The use of DNMT inhibitors such as decitabine and zebularine can inhibit DNMTs and subsequently activate methylation-silenced genes.17 The ANKRD11 mRNA expression was shown to be up-regulated in breast cell lines after treatment with different concentrations of these inhibitors, with zebularine showing the maximal response. This suggests DNA methylation as a likely mechanism for down-regulation of ANKRD11 expression. Following identification of the ANKRD11 promoter, we found a region located between –689 and +107 bp has enhanced ANKRD11 transcription. Analysis of this region in breast tumour samples identified three CpGs within the region –582 to –574 bp that were hypermethylated, however, it was not significantly related to ANKRD11 expression. In the in vitro methylation study, the methylation at specific CpGs of the ANKRD11 promoter was shown to reduce the transcriptional activity of ANKRD11, especially methylation at the third CpG. This supports a relationship of specific methylation and transcriptional activity of ANKRD11. However, site-directed mutagenesis of the third CpG did not affect the transcriptional activity, suggesting methylation and mutation of this region could be independent mechanisms in regulating the transcription of ANKRD11. The site-directed mutagenesis studies defined a region containing the two CpG dinucleotides as critical to the methylation-sensitive regulation of ANKRD11 transcription. In silico analysis identified GLI1 as a possible transcription factor that binds to this sequence 50 -GACCgCCCcg-30 ,18 but additional functional analysis is required to confirm such a relationship. The overexpression of DNMTs has been proposed as a mechanism for aberrant genome methylation.15 There was little variation in DNMT1 levels among our breast tumour samples. However, approximate 30% of tumours showed high levels of DNMT3B expression. In the methylation analysis, there was no significant relationship between specific methylation and ANKRD11 expression in human tumour samples, although there was a trend for low ANKRD11 expression and specific methylation of ANKRD11 in breast cancer cell lines. These findings speculate that mechanisms other than promoter methylation may be responsible for low ANKRD11 expression in some breast tumours. By comparing clinical data with specific methylation of ANKRD11, the pattern or methylation status of ANKRD11 promoter region PF1 was shown to be associated with HER2 status. As silencing of tumour suppressor genes through promoter hypermethylation is known to be a frequent and early event in carcinogenesis,19

3308

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309 HhaI

A aggcccaggggctctggagccgcggccgcgacgcgcccgaggg HhaI

ccgttctcttcgacggcagagcaagagtcaccgcgcgggagcc -582

HpaII -574

ccgagcgACCGCCCCGCTGCCCCGGTaccctcccctcgcagcc CG: M.SssI

gcccagagctcttc P6 MsssI.P6 MsssI.vector

**

*

HhaI.P6 HhaI.vector HpaII.P6

0. 20

0. 15

0. 10

0. 05

0. 00

HpaII.vector

Relative Luciferase Activity

B Site-mutagenesis: P6 CpG CpG CpG

(wild-type) triple mutant double mutant 3rd mutant

gACCGCCCCGCTGCCCCGGT gACAACCCAACTGCCCAAGT AAAAAAAAAGCTGCCCCGGT gACCGCCCCGCTGCCCAAGT

***

P6 **

P6-Triple mutant

n/s

P6-Double mutant P6-3rd mutant

0. 20

0. 15

0. 10

0. 05

0. 00

Empty vector

Relative Luciferase Activity

Fig. 5. In vitro methylation and site-directed mutagenesis of ANKRD11 promoter. (A) In vitro methylation target sequence and the luciferase activity of ANKRD11 promoter (pGL3-P6) located from –689 to –543 bp upstream of the transcription start site. (B) Site-directed mutagenesis of specific sequences and luciferase activity of pGL3-P6. Unpaired two-tailed t-tests were conducted on each construct compared to the original P6. Recognition sequences for HhaI (50 -GCGC-30 ) and HpaII (50 -CCGG-30 ) are shown in bold. Recognition sequences for M.SssI (50 -CG-30 ) are underlined. Unit 7 is shown in uppercase text. *p < 0.05, **p < 0.005, ***p < 0.001, n/s, not significant.

methylation status of ANKRD11 could be a candidate marker for early tumour detection. In conclusion, this study is the first to identify ANKRD11 promoter hypermethylation in breast tumours. The frequently observed hypermethylation of the ANKRD11 promoter and down-regulation of ANK RD11 mRNA levels in breast tumours point towards an association between the inactivation of this tumour suppressor gene and aberrant breast growth.

Acknowledgements

Conflict of interest statement

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/ 10.1016/j.ejca.2012.03.023.

None declared.

This work has been supported by the Cancer Agency, the National Health & Research Council, Australia, and the Government’s Operational Infrastructure Program.

Victorian Medical Victorian Support

Appendix A. Supplementary data

S.P. Lim et al. / European Journal of Cancer 48 (2012) 3300–3309

References 1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA A Cancer J Clin 2011;61:69–90. 2. Addou-Klouche L, Adelaide J, Finetti P, et al. Loss, mutation and deregulation of L3MBTL4 in breast cancers. Mol Cancer 2010;9:213. 3. Neilsen PM, Cheney KM, Li C-W, et al. Identification of ANKRD11 as a p53 coactivator. J Cell Sci 2008;121:3541–52. 4. Zhang A, Yeung PL, Li C-W, et al. Identification of a novel family of ankyrin repeats containing cofactors for p160 nuclear receptor coactivators. J Biol Chem 2004;279:33799–805. 5. Isrie M, Hendriks Y, Gielissen N, et al. Haploinsufficiency of ANKRD11 causes mild cognitive impairment, short stature and minor dysmorphisms. Eur J Hum Genet 2011;20:131–3. 6. Willemsen MH, Fernandez BA, Bacino CA, et al. Identification of ANKRD11 and ZNF778 as candidate genes for autism and variable cognitive impairment in the novel 16q24.3 microdeletion syndrome. Eur J Hum Genet 2010;18:429–35. 7. Callen DF, Crawford J, Derwas C, et al. Defining regions of loss of heterozygosity of 16q in breast cancer cell lines. Cancer Genet Cytogenet 2002;133:76–82. 8. Katoh H, Shibata T, Kokubu A, et al. Epigenetic instability and chromosomal instability in hepatocellular carcinoma. Am J Pathol 2006;168:1375–84. 9. Brenet F, Moh M, Funk P, et al. DNA methylation of the first exon is tightly linked to transcriptional silencing. PLoS One 2011;6:e14524. 10. Harder A, Titze S, Herbst L, et al. Monozygotic twins with neurofibromatosis type 1 (NF1) display differences in methylation

11.

12.

13.

14. 15.

16.

17. 18.

19.

3309

of NF1 gene promoter elements, 50 untranslated region, exon and intron 1. Twin Res Hum Genet 2010;13:582–94. Wong NC, Novakovic B, Weinrich B, et al. Methylation of the adenomatous polyposis coli (APC) gene in human placenta and hypermethylation in choriocarcinoma cells. Cancer Lett 2008;268:56–62. Lim SP, Neilsen P, Kumar R, Abell A, Callen DF. The application of delivery systems for DNA methyltransferase inhibitors. BioDrugs 2011;25:227–42. Chen T, Ueda Y, Dodge JE, Wang Z, Li E. Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b. Mol Cell Biol 2003;23:5594–605. Bestor T. The DNA methyltransferases of mammals. Hum Mol Genet 2000;9:2395–402. Girault I, Tozlu SL, Lidereau R, Bieche I. Expression analysis of DNA methyltransferases 1, 3A, and 3B in sporadic breast carcinomas. Clin Cancer Res 2003;9:4415–22. Julia Grabowski M, Sidney L, Salzstein M, Georgia Robins Sadler M, Sarah Blair M. Intracystic papillary carcinoma: a review of 917 cases. Cancer 2008;113:916–20. Yoo CB, Jones PA. Epigenetic therapy of cancer: past, present and future. Nat Rev Drug Discov 2006;5:37–50. Ji H, Vokes SA, Wong WH. A comparative analysis of genomewide chromatin immunoprecipitation data for mammalian transcription factors. Nucleic Acids Res 2006;34:e146. Brooks J, Cairns P, Zeleniuch-Jacquotte A. Promoter methylation and the detection of breast cancer. Cancer Causes Control 2009;20:1539–50.